Theory of Knowledge

Theory of Knowledge with Kurt Sylvan

Syllabus

Course Description

Epistemology is a major branch of philosophy dedicated to questions about the nature and structure of knowledge, justified belief, evidence, good inference, etc. Examples of central questions include:

  • What are the necessary and sufficient conditions for knowledge?
  • To gain knowledge from a reliable source, does one also need to know that the source is reliable?
  • Must all beliefs be justified by other beliefs, or are there some non-inferentially justified beliefs?
  • Are the factors that make for justified belief all internal, or are some external?
  • Is evidence necessary for justified belief? If so, is good evidence all one needs for justification?
  • Do we know anything? Can we answer the skeptic? (Must we answer the skeptic?)

We will read and argue about some classic answers to these questions (and others) in the class.

Readings

Many readings will come from Epistemology: An Anthology (eds. Sosa, Kim, Fantl and McGrath). I also recommend (but don’t require) reading from Richard Fumerton’s introductory book Epistemology. These books and almost everything else that we will read are available on the Sakai site for the class. A few background readings will also be drawn from the Stanford Encyclopedia of Philosophy, which is free online.

Here is the order in which we will cover the topics, and what we will probably read (broken down into required readings and recommended readings):

  • 5/28: The Justified True Belief (JTB) Analysis of Knowledge and the Gettier Problem
    • Required. Gettier, E. “Is Justified True Belief Knowledge?” in E:AA
    • Recommended. Ichikawa, J. & Steup, M. “The Analysis of Knowledge”
    • The Analysis of Knowledge
  • 5/30: Responses: Defeasibility Theories and Causal Theories of Knowledge
    • Required. Harman, G. Selections from Thought in E:AA
    • Goldman, A. “A Causal Theory of Knowing” separately on Sakai
    • Recommended. Lehrer and Paxson. “Knowledge: Undefeated Justified True Belief.” on Sakai Klein, P. “Knowledge, Causality and Defeasibility.” on Sakai
    • Sudduth, M. “Defeaters in Epistemology”
    • Defeasibility Analyses and Propositional Defeaters
  • 6/4: Responses: Sensitivity and Safety Theories
    • Required. Nozick, R. “Knowledge and Skepticism” in E:AA
    • Sosa, E. “How to Defeat Opposition to Moore” in E:AA
    • Recommended. Comesaña, J. “Knowledge and Subjunctive Conditionals” on Sakai
  • 6/6: Responses: Virtue Epistemology a la Sosa
    • Required. Sosa, E. Apt Belief and Reflective Knowledge, v.1 Chapters 2 and 5 on Sakai
    • Recommended. Greco, J. and Turri, J. “Virtue Epistemology”
    • Stanford Encyclopedia of Philosophy
  • 6/11: The Structure of Justification
    • Required. BonJour, L. “Can Empirical Knowledge Have a Foundation?” in E:AA
    • Alston W. “What’s Wrong with Immediate Knowledge?” separately on Sakai
    • Recommended. Fumerton, R. “Foundationalist Theories of Epistemic Justification”
    • Stanford Encyclopedia of Philosophy
    • Klein, P. “Human Knowledge and the Infinite Regress of Reasons.” in E: AA Alston, W. “Level Confusions in Epistemology.”
  • 6/13: Internalism and Externalism
    • Required. Fumerton, R. “Traditional (Internalist) Foundationalism” (Ch. 4 of Epistemology) Goldman, A. “What Is Justified Belief?” in E:AA
    • Recommended. Fumerton, R. “Externalist Versions of Foundationalism” (Ch. 4 of Epistemology)
  • 6/18: Internalism and Externalism
    • Required. Goldman, A. “Internalism Exposed” in E:AA Vogel, J. “Reliabilism Leveled” in E:AA
  • 6/20: Evidentialism and Reasons in Epistemology
    • Required. Conee and Feldman. “Evidentialism” in E:AA
    • Sylvan, K. “Reasons and the Metaphysics of Justification.”
  • 6/25: The Problem of Easy Knowledge
  • 6/27: Skepticism and Closure
    • Required. Dretske, F. “Epistemic Operators” in E:AA
    • Stine, G. “Skepticism, Relevant Alternatives, and Deductive Closure” in E:AA
    • Recommended. Weatherson, B. “Sceptical Arguments and Sceptical Scenarios” First part of: Skepticism notes
  • 7/2: Skepticism and Underdetermination
    • Required. Brueckner, A. “The Structure of the Skeptical Argument” on Sakai Pritchard, D. “The Structure of Sceptical Arguments.” on Sakai
  • 7/4: Dogmatism and the Modest Anti-Skeptical Project
    • Required. Pryor, J. “The Skeptic and the Dogmatist.” on Sakai Recommended. Pryor, J. “What’s Wrong with Moore’s Argument?”

Assignments, Grading and Policies

Your grade will be based on:

  1. Discussion Board Comments/Questions. You are responsible for posting
    1. some questions about the reading or (II) some critical comments or objections to the arguments in the reading on the Discussion Board on Sakai. These can be anywhere from a few sentences to several pages what matters is that they help you learn the material. You should post these, at latest, at 9:00 on the night before the class meeting. Each of you should do this for 5 meetings of your choosing. This is for 20% of the grade.
  2. Attendance and Participation. You must the class regularly to do well and avoid getting lost. I will allow one unexcused absence, and you will be penalized for every subsequent unexcused absence. You are also encouraged to participate by talking as much as possible. This will be for 10% of the grade.
  3. Exam. There will be a take-home mid-term exam. I’ll hand out questions on 6/13 and you will have a week to write answers to them, and hand them into me in printed form on 6/20. This is for 30% of the grade.
  4. Final Paper. A final paper of 6-10 pages is due on 7/4 or earlier by email. I’ll post topics on 6/13. You can also design your own topic, but you must pass it by me first. This is for 40% of the grade. Do not plagiarize on your papers! This is severely penalized by the University and easy to detect.

Contact Information and Office Hours

You can contact me via ksylvan@philosophy.rutgers.edu. I will hold office hours in Seminary 1, Room 012 on Tuesdays and Thursdays from 4:00-5:30. Feel free to arrange an appointment with me if you can’t make the office hours and have any questions or want to discuss your work.

May 28th, 2013 Reading

“Is Justified True Belief Knowledge?” Edmund Gettier

  • What are the necessary and sufficient conditions for knowledge?
    • Those attempts are of a form similar to:
    1. S knows that P iff
    2. P is true.
    3. S believes that P, and
    4. S is justified in believing that P.
  • It is possible in being justified in believing something that is in fact false.
    • For any proposition P that S is justified in believing, where P entails Q and P accepts Q, then S is justified in beliving Q.

Case I

  1. Jones is the man who will get the job, and Jones has ten conns in his pocket.
  2. The man who will get the job has ten coins in his pocket.
  • Smith has strong evidence for (d).
    1. entails (e).
  • Smith accepts (e) on the grounds of (d), which he is justified in doing.
  • But unknown to Smith, he, not Jones, will get the job.
  • He also happens to have 10 coins in his pocket.
    • So (e) is true.
    • Smith believes (e).
    • Smith is justified in believing (e).

Case II

  1. Jones owns a Ford.
  2. Either Jones owns a Ford, or Brown is in Boston.
  3. Either Jones owns a Ford, or Brown is in Barcelona.
  4. Either Jones owns a Ford, or Brown is in Brest-Litovsk.
  • Smith has strong evidence to believe that (f).
  • Smith accepts (g), (h), and (i) on the basis of (f).
    • Smith is completely justified in doing so.
  • Imagine that Jones does not own a Ford, but is driving a rented car.
    • And by coincidence, Brown happens to be in Barcelona, making
      1. true.
  • Therefore,
      1. is true.
    2. Smith does believe that (h) is true.
    3. Smith is justified in believing that (h) is true.

Conclusion

    1. does not state sufficient conditions for knowledge.

“The Analysis of Knowledge” Ichikawa, J. & Steup, M.

The Analysis of Knowledge

May 28th, 2013 Handout

The Search for the Tether and the JTB Theory

Plato’s Tether

Our question for the next several meetings:

  1. What distinguishes knowledge from a lucky guess (or a mere true opinion)?

(Q) goes back to the Meno, where Plato puts a metaphorical answer in Socrates’ mouth:

SOCRATES: True opinions, for as long as they remain, are fine things and do nothing but good. But they don’t hang around for long; they escape from a man’s mind, so that they are not worth much until one tethers them with chains of reasons why. […] That is why knowledge is prized more highly than correct opinion; knowledge differs from correct opinion in being tethered down.

MENO: Yes, by Zeus, Socrates, it does seem to go something like that.

JTB

But what is this tether? Socrates suggests that it consists in a chain of reasons why. A natural way to make this more precise is to think that what distinguishes K from mere TB is justification: someone who makes more than a lucky guess is justified in believing as she does. Exactly what justification involves is a huge question. For the moment, let’s just use the intuitive idea, and consider this answer to (Q):

(JTB) Justification is what distinguishes knowledge from a lucky guess.

In other words: knowledge  =  justified true belief.

After this suggestion hit the philosophical newsstand, little changed for the ensuing 2300+ years. Even in the first half of the 20th century, JTB was accepted in some form or another. One can find versions, e.g., in Chisholm and Ayer:

(Ayer’s view) S knows that P iff (i) S is sure that P, (ii) S has the right to be sure that P, and (iii) P is true.

(Chisholm’s view) S knows that P iff (i) S accepts P, (ii) S has adequate evidence for P, and (iii) P is true.

Elements of JTB

Before moving on, it’s worth saying a bit about the elements of the JTB theory, since it isn’t wholly uncontroversial that each is really necessary for knowledge:

  • Belief: It’s plausible that one can’t know some proposition without at least believing it. This is witnessed by the absurdity of sentences like: ‘I know that it’s raining but I am not sure that it’s raining’. This simply sounds incoherent. Nonetheless, there are some cases where one can put pressure on the belief condition. Consider:
  • Exam: Albert is quizzed on English history. One question is: ‘When did Queen Elizabeth die?’ Albert doesn’t think he knows, but answers the question correctly, by writing that Elizabeth died in
    1. Moreover, he gives correct answers to many other questions to which he didn’t think he knew the answer.
    • There’s some temptation to say that Albert knows that Elizabeth died in 1603 but doesn’t believe that she died in 1603.
    • Still, we need to distinguish between implicit and explicit belief. Consider someone who has a prejudice against some group of people and is disposed to treat them as inferior. This person may explicitly deny that he has this prejudice, and not believe that he has it. Still, given his behavior, it seems appropriate to ascribe him as having the implicit belief that members of this group are inferior. Given the distinction between implicit and explicit belief, it is also plausible that Albert does at least implicitly believe that Elizabeth died in 1603.
  • Truth: Sometimes we say things like ‘I just knew it was going to happen. But then it didn’t!’ These claims don’t sound incoherent. But wouldn’t they be incoherent if knowing that P entailed P? The epistemologist’s response is to allow that the English word ‘knows’ might be used non-literally to express the speaker’s feeling of knowing.
    • Still, once we are engaged in philosophical elucidation, it is more helpful to distinguish between the mere feeling of knowing and actual knowing (assuming there is some actual knowing, pace radical skeptics). Perhaps the word can express both. But as long as we can conceptually distinguish the two we can simply recognize that the JTB theory isn’t about the feeling of knowing but about actual knowing.
  • Justification: A large can of worms. We’ll see that some major epistemologists have denied that justification is really necessary for knowledge. One can get the feeling for this by simply thinking again about exam. If Albert does know that Elizabeth died in 1603, what justification could he possibly have for believing that she died in 1603? It sure seems that he doesn’t have evidence for believing this.
    • The main response to this challenge is to insist that being justified in believing something needn’t consist in having evidence for believing something, though evidence is one way to get justified. Perhaps, for example, reliable sources like memory and intuition, etc., can yield justified beliefs that don’t require independent evidence.

Gettier: JTB ≠ K

Gettier’s little paper presents two simple counterexamples to JTB. As we’ll be seeing, there are heaps of similar cases that pop up to undermine revisions of JTB (i.e., ‘JTB+’ views).

(Case I) Smith has strong evidence for the following proposition:

  1. Jones will get the job and has ten coins in his pocket.

This entails:

  1. Someone who has ten coins in his pocket will get the job. By coincidence, Smith rather than Jones will get the job. Smith also has ten coins in his pocket, though he doesn’t realize it. So, (Q) is true and Smith is justified in believing (Q).

But Smith does not know (Q).

(Case II) Smith has strong evidence for the following proposition:

  1. Jones owns a Ford.

This entails

  1. Jones owns a Ford or Brown is in Boston.

Smith deductively infers (S) from (R), and so justifiedly believes (S). By chance, Brown is in Boston, and (S) is true only for this reason.

But Smith does not know (S).

There are more intuitive cases that one could give that have this structure. Consider:

(Sheep in the Field) Smith goes for a drive in the country. He looks off into a field and sees what looks exactly like a sheep. So he justifiably believes:

That animal in the field is a sheep.

Smith’s son is in the back seat reading a book and not looking at the scenery. He asks if there are any sheep in the field they are passing. Smith says ‘Yes’, adding:

There is a sheep in the field.

Smith is justified by what he sees in believing (1). (2) follows from (1), so he is justified in believing (2). As it turns out, (1) is false: what Smith sees is a sheep dog. But (2) is true anyway: there’s a real sheep just behind the sheep dog which Smith can’t see.

A Cheap Response: No False Grounds

Gettier gives us the recipe he used for generating these particular cases in the paper. Cases I and II are simply cases where (i) one is justified in believing something false but (ii) infers something true from it using a good method. The inferred true belief will be justified given:

(Deductive J-Closure) For any proposition P, if S is justified in believing P, P entails Q, and S deduces Q from P and accepts Q as a result of this deduction, then S is justified in believing Q.

Seeing this, some people in the early literature thought it was easy to revise JTB to avoid Gettier’s problem. They suggested:

(JTB+No False Grounds) S knows that P iff (i) S believes that P, (ii) P is true, (iii) S is justified in believing P, and (iv) all of S’s grounds for believing P are true.

But this proposal faces two problems: it is too weak and too strong. First, it is easy to construct cases where the subject doesn’t reason from any false premises. Consider:

Henry drives into a part of the state that he’s never visited before. He is looking at a real barn, and has impeccable visual and other evidence that it is a barn. His justification is sound in every way. However, in the rest of this unusual part of the state there only fake, papiere-mache barns, any of which would have fooled Henry into thinking it was a barn.

Here, Henry doesn’t engage in any inference at all: he just looks at the barn and believes that there’s a barn there. He does not reason from anything false. Still, he doesn’t know. For it is just by sheer luck that he finds himself in front of the one real barn in the area.

Another reason why JTB+No False Grounds fails is that it’s too strong. We can sometimes know things even though some of our grounds for belief are false. Consider:

Smith has two independent sets of reasons for thinking that someone in his office owns a Ford. One set of reasons has to do with the wily Nogot, who says that he owns a Ford, drives around a Ford (which isn’t really his, unbeknownst to Smith), etc. But Smith has equally strong reasons having to do with Havit. And Havit does own a Ford.

In this kind of case, Smith might reason to the true conclusion that someone in his office owns a Ford from the whole set of reasons, some of which are false. Intuitively, Smith can still know that someone in his office owns a Ford even though some of these reasons are false. But JTB+No False Grounds predicts otherwise. So JTB+No False Grounds fails.

Epistemic Luck

If one wants to solve the Gettier problem while holding onto a broadly JTB+ theory, the strategy will be to (i) find an element that all Gettier cases have in common and (ii) add to JTB a clause that says that this element must be absent. We’ve seen one unpromising way to do thisthe ‘no false grounds’ view. But in seeing why this view failed, we got a hint of a more abstract feature that these cases share. Remember what we said about Henry in fake barn country: ‘It is just by sheer luck that he finds himself in front of the one real barn in the area’.

This does seem to be an important feature of the cases we’ve seen so far. In all of them, the person lands on a true belief only by luck. It was a sheer coincidence in Case I that Smith himself also had 10 coins in his pocket. It was a sheer coincidence in Case II that Smith happened to infer a disjunction whose second disjunct (i.e., that Brown is in Boston) was true. He could so easily have inferred a different disjunction whose second disjunct was false. It was a sheer coincidence in Sheep in the Field that there was a real sheep hiding behind the sheepdog. Etc.

This isn’t to say that these people engaged in lucky guessing. They were justified in believing what they believed. So, there’s a different kind of luck that arises in Gettier cases.

Now, we can’t just say that no kind of luckily true belief is compatible with knowledge. Consider some other examples where knowledge isn’t missing.

  • It’s in some sense a matter of luck that I’m in this room. The last five years could so easily have been different, if slight changes were made to the years before them. But that doesn’t mean that I don’t know that there are students in this room.
  • Furthermore, if I look out the window and happen to see something really rare and bizarre, it hardly follows that I can’t know that this rare and bizarre thing is happening. Still, there’s a sense in which the beliefs I form are true by luck.
  • Another example from Ichikawa and Steup: ‘Suppose that someone enters a raffle and wins an encyclopedia, then reads various of its entries, correcting many of his previous misapprehensions. There is a sense in which the resultant beliefs are true only by luckfor our subject was very lucky to have won that rafflebut this is not the sort of luck, intuitively, that interferes with the possession of knowledge.’

So, how should we understand the luck that matters? This question is as hard to answer as the Gettier problem is to solve. So we can’t solve the Gettier problem with a theory like:

  • (JTB+No Luck) S knows that P iff (i) S believes that P, (ii) P is true, (iii) S is justified in believing P, and (iv) S doesn’t believe truly merely by luck.

This simply provides a new label for the problem without solving it.

Still, it does point us in a helpful direction. If we could provide an analysis of the relevant notion of a belief’s being true merely by luck, we might stand a chance of solving the Gettier problem. So we’ve made some progress: we’ve ruled out one unhelpful approach and have gotten a grip on a strategy that will guide us. In future meetings, we’ll see some proposals that might help us to understand the nature of epistemic luck.

May 28th, 2013 Lecture

Gettier

  • Some people try to make a strong connection between knowledge and justification.
    • Since Gettier, people have dropped this for “JTB+ view”
  • The story is that Gettier was up for tenure review, and his publication history wasn’t great.
    • So he quickly cooked up this paper “Is Justified True Belief Knowledge?”
  • The article consists of 2 counterexamples about the belief that knowledge contains justified true beliefs (JTB).

Epistemic Luck

  • There are some scientific theories that assume things that just cannot be true.
    • Like Newton’s theory of gravitation, not as good as relativity.
    • You can come to know true things from false premises.

May 30th, 2013 Reading

Selections from Thought in E:AA by Gilbert Harman

Knowledge and Probability

The lottery paradox
  • Some people think we never believe anything for which we are not certain.
    • We have higher and lower “subjective probability.”
    • If knowledge implies belief, we never know anything that isn’t absolutely certain.
    • This conflicts with intuitions about knowledge, where we think we know some things with a greater probability than we think we know others.
  • We might count as belief anything whose “subjective probability” exceeds . 99.
    • This would also conflict with ordinary views.
    • We do not suppose that a man inconsistently believes of every participant in a fair lottery that the participant will lose.
      • Even if we suppose that the man assigns a subjective probability greater than . 99 to each’s person’s losing.
    • If ordinary views are to be preserved, belief must be distinguish from high degree of belief.
  • A rule of inductive inference is sometimes called a “rule of acceptance”
    • It tells us what we can accept given other beliefs and degrees of belief.
    • A purely probability rule of acceptance says that we may accept something if and only if its probability is greater than . 99.
    • Kyburg points out that such rule leads to the “lottery paradox” because it authorizes an inconsistent set of beliefs.
  • It is paradoxical to suppose that we could rationally believe of every participant in a lottery that he will lose.
Gettier examples and probabilistic rules of acceptance
  1. Reasoning that essentially involves false conclusions, intermediate or final, cannot give one knowledge.
  • Probabilistic rules of acceptance do not permit an explanation of Gettier examples by means of (P).
  • A defender of purely probabilist rules might reply that in a Gettier case, the believer of the false “knowledge” must infer her conclusion not that she must conclude not from the belief but from the evidence.
    • But given any evidence, some false conclusion will be highly probable.
    • For example, let (s) be a conclusion saying under what conditionals the NJ State Lottery was most recently held.
      • Let (q) say what ticket won the grand prize.
      • Consider the conclusion “not both (s) and (q)” and call this (r).
      • The conclusions (r) is highly probable, but (r) is false.
      • If such highly probable false conclusions were always considered essential to an inference, the Gettier believer could never come to know anything.
  • The problem is purely probabilistic considerations do not suffice to account for the peculiar relevance about the ownership of the car.
  • The trouble is that purely probabilistic rules are incompatible with the natural account of Gettier examples by means of principle (P).
    • The solution is not to modify (P) but modify inference.

Knowledge and Explanation

A causal theory
  • Goldman suggests we know only if there is the proper sort of causal connection between our belief and what we know.
    • For example, perceiving something and believing that something is there is somehow properly causally protected.
  • The Gettier believer fails because the causal connection is lacking.
  • General knowledge does not fit into this simple framework.
    • All emeralds being green is neither caused nor caused by the particular green emeralds we come to know.
  • Perhaps a person knows by inference only if all conclusions essential to that inference are true.
    • The inference must satisfy (P).
    • Mary fails to know because she is wrong when she infers that Nogot’s past ownership is responsible for Nogot’s present ownership.
Inference to the best explanatory statement
  • It gets better is we replace “cause” with “because.”
  • This account better explains ordinary usage.
    • Nogots past ownership explains Mary’s evidence, but it is odd to think it is caused by Mary’s evidence.
    • The detective infers that activities of the butler explain these footprints, does he infer these activities caused these footprints?
  • Moving from “cause” to “because” avoids Goldman’s ad hoc treatment of knowledge.
  • “X’s tend to be Y’s will explain the next X will be Y” is sufficiently more plausible that competitors such as “interfering factor Q will prevent the next X from being a Y.”
    • The doctor says you will get measles.
    • Because doctors are normally right, it will explain why the doctor is right in this case.
Further examples
  • There is a box of Sure-Fire matches.
    • I infer that the next match will late on the premise that all the others have lit.
    • It turns out the next match is a defective match, that needs to be raised to 600 degree to light.
    • I strike the match and a sudden blast of Q-radiation to 600 degrees, and the match lights. Did I know it would light? No.
Statistical inference
  • You walk into a casino and see the roulette wheel stop at red fifty times in a row.
    • The explanation may be that the wheel is fixed.
      • It could be that the wheel is fair and this is one of those times when fifty reds comes up in a fair wheel.
    • But if the explanation is that the wheel is fair and that this is just one of the times, it says what the sequence of red is the result of, the “outcome of.”
      • It does not say why fifty red occurred this time rather than another time.
Conclusion
  • We are led to construe induction as inference to the best explanation.
    • Inductive reasoning is seen to consist in a sequence of explanatory conclusions.
  • We are led here because Gettier cases need to be accounted for.
  • The present model replaces Goldman’s “causal” with “explanatory.”

Evidence One Does Not Possess

Evidence against what one knows

One knows if there is no evidence such that if one knew about the evidence one would be justified in believing one’s conclusions.

“A Causal Theory of Knowing” by Alvin Goldman

  • Begins with Gettier.
    1. Jones owns a Ford.
    2. Either Jones owns a Ford or Brown is in Barcelona. (“Or-in”)
  • One hypothesis for solving this is to say that Smith cannot be said to know this because it’s false, and so he can’t justify something from it.
    • What Goldman will do is take this specific case, analyze it, and generalize to a new take on “S knows that p.”
  • What makes (p) true is the fact that Brown is in Barcelona, but that has nothing to with Smith’s believing that (p).
    • Alternatively, if Jones owned a Ford, and his owning a Ford manifested by his offer of a ride to Smith, and he offered Smith a ride, we’d say that Smith knew Ford.
    • The requirement of such a causal connection is what Goldman wants to add.
  • One of the good examples of an appropriate causal connection is that of perception (specifically the one defended by H. P. Grice)
    • Where (S) sees that there is a vase in front of him.
      • A necessary condition of (S)’s seeing that there is a vase is that there be a certain kind of causal connection between the presence of the vase and (S)’s believing that a vase is present.
    • Largely, this is a “problem” for special science and not philosophy.
  • Goldman comes up with veritable and deviant causal chains.
    • There, evidently, needs to be a “special sort,” where the presence of the vase is connected with (S)’s belief.
  • What of the scope of perceptual knowledge?
    • I can know of the vase.
    • Can I know a painting is from Picasso?
    • “Left indeterminate.”
  • Now consider memory.
    • A causal connection between earlier belief of (p) and later belief of (p) is certainly a necessary ingredient in memory.
      • This is the same as his last argument, maintained over t0 to t1.
  • He can connect perception and memory, but more interestingly, Goldman moves to inference.
  • Infernece is to say that (S) knows that (P) and does not entail an explicit, consicious process of reasoning.
  • Suppose (S) perceives that there is solidified lava in the countryside.
    • Where (p) and (q) are the propositions, 
      (p) ---> (q) ---> B_s(q)\
                         \
                          \
                B_s(r) ---> B_s(p)
    1. can come to know (p) when he correctly reconstrucs the causal chain leading from (p) to the evidence for (p) that (S) perceives.
    • What’s illusory about this, I think, is “correctly.” I think this is just a “step up” from saying that (S) is justified in (p).
  • “Appropriate” knowledge producing causal processes include:
    1. Perception
    2. Memory
    3. A causal chain, pattern 1 or 2, correctly reconstructed by inferences.
    4. Combinations of (1), (2), and (3).
  • Meta-analysis of knowledge is not causal.

May 30th, 2013 Lecture

  • We looked at the original Gettier in the last class and saw how JTB is not knowledge.
    • Perhaps no false assumptions?
    • Today we’ll see a more sophisticated version of this.

Defeasibility Theories and Objective Defeaters

  • Before Harman, there is another tradition that Harman criticizes.
    • The theory was the “defeasibility theory.”
  • On one hand, we talk about evidence in an objective sense, and we search for facts.
    • There’s objective evidence, evidence which is “out there,” but they are facts that we do not have access to.
    • Then possessed evidence, which is what we actually have.
  • With this there are two kinds of normative statuses:
    • A correct belief is one that is supported by all objective evidence that exists.
    • An epistemically rational belief is one supported by the evidence we have.
  • There are two types of defeaters, for two kinds of defeat:
    • A rebutting defeater is a counter-example, or evidence to suggest the opposition.
    • A undercutting defeater undermines your evidence, like learning that your source was faulty. It says nothing about the fact of the matter.
  • All of people had a stipulative definition of an objective defeater that does not lead to a “true/undefeated” proposition.
    • The stated objections rely on the definition.
  • Objective defeaters are the evidence you do not have in a Gettier case.

Harman’s Theory

  • He provides a more sophisticated version, the reasons why it’s more sophisticated:
    • He adds the condition that it has to be the case that you don’t essentially rely on any false ground.
      • If you know you must have no false grounds, which predicts the wrong things.
    • Widen the notion of inferences, and allow for them to all be knowledge, there are these cases that look like non-inferential cases.
      • In the fake barns case, you’re relying on the assumption that conditions are normal, that appearance reliably correspondends to reality.

Goldman’s Theory

  • According to Goldman, what’s going on in Gettier cases is causal.
    • The holding a belief and the facts are not casually related, and that’s what’s wrong.
  • You have knowledge if and only if you have there is an appropriate causal relationship between the justification and the truth/knowledge.
    • Where is the “appropriateness.”
  • There are “deviant causal chains” where things are wholly related to the facts in an intuitively non-perceptual, false-memory, etc, way.

June 4th, 2013 Reading

“Knowledge and Skepticism” Robert Nozick

Knowledge

  • Our task is to formulate further conditions to go alongside:
  1. p is true.
  2. S believes that p.
  • We want:
    • Each condition to be necessary.
    • Jointly sufficient for knowlege.
  • The causal condition on knowledge is inhospitiable to mathematical and ethical knowledge.
  • Consider the third condition (two formulations):
  1. If p weren’t true, S woudn’t believe that p.
  2. ¬p → ¬B(S, p)

“How to Defeat Opposition to Moore” by Ernest Sosa

“Knowledge and Subjunctive Conditionals” by J. Comesaña

June 4th, 2013 Handout

June 4th, 2013 Lecture

  • Now we’re going to talk about theories that hold that it’s just an accident that people form JTBs.
  • It’s reasonable if you think about someone infering from Jones owning a Ford to something false, and then you notice that while it’s justified, it is an accident.
    • People want to give an account of the accident.
    • Non-accidentality

The Tether in Modal Terms

  • When philosophers talk about modality, they are talking about statements like “could”, “would” “must.”
  • The counterfactual or subjenctive conditional statement is of the form:

If it were true that A, it would be true that C.

  • Philosophers have used “nearby possible worlds” as a concept recently.

In similar possible circumstances when A, then C is also true.

(Sensitivity) S knows that P only if if P were false, S would not believe that P.

(Safety) S knows that P only if S would believe P only if P were true.

  • These are distinct logical conditions
  • Order in questions:
    • First order questions are like “does P know that S?” *K**p*
    • Second order questions are “does P know that P knows S?” KKp
  • A good analogy is like thermometers: A thermometer that says 70 degrees but is broken is not right when it is actually 70 degrees.

Nozick and Sensitivity

  • Nozick notices that there is a counterexample which he calls the (Grandma) case.
    • When her grandson is doing well, she observes that he is doing well.
    • When he is not doing well, the family make her believe he is.
  • The way that he accounts for this counterexample is by saying you must always have the same method for coming to know something in both cases.
  • Relativizing to methods is a richer and more accurate way of thinking about it because you want to keep some variables constant.
  • I do not think that in lottery cases you can assert that your ticket will lose because you are presented with a precision of facts do not exist in other belief cases. Knowledge and Lotteries, book, read.
  • Virtues:
    1. Explains many Gettier cases.
    2. Explains you can “never know” if your ticket is a loser.
    3. Explains skeptics worries and yields the right response.

(The argument from ignorance):

  1. You know that: if you have hands, then you’re not a brain in a vat.
  2. (Closure) If you know that P and know that if P, then Q, then you know that Q.
  3. But you don’t know that you’re not a brain in a vat.
  4. So, you don’t know that you’re not a brain in a vat.
  • Vices
    1. (Closure) is right, making you wrong.

June 6th, 2013 Reading

Apt Belief and Reflective Knowledge Chapters 2 and 5, by Ernie Sosa

Lecture 2 A Virtue Empistemology

  • When an archer takes aim and shoots, that shot is assessable in three respects:
    1. Whether he succeeds in its aim, hitting the target.
    2. Whether it is adroit, manifests the skills of the archer.
    3. Whether it is apt, whether it is because he is accurate and adroit.

“Virtue Epistemology” by Greco, J. and Turri, J.

Analysis of Knowledge Stanford Encyclopedia of Philosophy

June 6th, 2013 Lecture

Nozick and sensitivity

  • Nozick came up with the sensitivity argument.
  • He noticed that it entailed it that closure is false.
  • Being clever, he also realized he could use it to respond to the skeptic.
    • This meant that it was a virtue and not a vice.
  • If you wanted to motivate sensitivity on it’s own, there’s a lot you can do.
    • Very often, we want sensitivity.
    • If your beliefs were false and you’d still hold them, that’d be a problem.

Safety

  • We can imagine easily a world where a common scenario is found in a dream, making you wrong, a wrong belief.
    • So dreams have to be fake.
  • Well you wake up a realize it was a wrong belief.
    • The person in fake barn country could learn more about their enviroment, and realize that they weren’t looking at a barn or that they weren’t allowed to use their assumption about the barn
  • “Retrospective knowledge” is dangerous.
  • Is this similar to the defeater problem?
    • The simpler thing to say is that we can still evaluate the prior time slice, while they later went on to reform their beliefs, in the moment they didn’t know.

An aside Helpful Wishful Thinking

  • There are some cases where a worry similar to circular reasoning arising.
  • Safety was proposed as a necessary condition on knowledge.
    • What do you have to add to JTB to get knowledge?
    • If you pursue safety for this, you get to avoid the helpful wishful thinking.

Truth fairy

  • Devastating for necessary truth.
  • There is something behind the circular reasoning thought.

Virtue epistemology

  • The more general pattern of evaluation is the AAA pattern.
  • On an analogy about archers,
    • If it hit the bull’s-eye, it is accurate.
    • If it manifest the archer’s competence, it is adroit.
    • If the success of a shot was a manifestation of the archer’s competence, it is apt.
  • The nice thing about Sosa’s picture is that it allows us to understand Gettier cases as a specific instance of a failure of aptness.

First order and second order seperation motivation

  • We don’t need a criterion of truth to act.
    • Does this mean you can act on belief?
  • From an abstract point of view, it is better to act on knowledge rather than abstract belief.

Sosa and Analysis of knowledge

  • Sheep in a field
    • The accurary of the belief is not determined by your aptness.
    • There is not apt belief.
  • You’ve bought a box of surefire matches.
    • There’s only one which doesn’t reliably light.
    • But when you strike it, a peice of Q-radiation lights it.

Animal knowledge and reflective knowledge

  • Animal knowledge is knowledge of primitive things, first order.
    • It is apt belief.
  • Reflective knowledge is “apt belief aptly noted.” RKp = KKp
  • add something

June 11th, 2013 Reading

Is there Immediate Justification? by James Pryor

Justification

  • Can justification ever happen “immediately” or “non-inferentially”?
  • There could be some mix up about what is an isn’t justification with the author.
  • You have justification to believe *P** if and only if you are in a position where *P is epistemically likely for you to be true.
    • A very inclusive epistemic status.
  • All it means for Pryor is that it is appropriate for you to believe it.
  • If there is some state or condition you are in in virtue of which you have justification to believe P, Pryor will call is “justification making condition.”
  • It will be useful for us to distinguish between having justification to believe and appropriately holding a belief.
    • To have justification, it is not important whether you actually believe.
      • There just have to be things that make believing an appropriate attitude.
    • Appropriate belief requires more, you need to believe it, you need to have justification, and you need to believe on the right grounds.
      • You need to believe it for the reasons that make you have justification to believe it.

Immediate Justification

  • There are some propositions that are believed because of other prior justification epistemically supporting them.
    • Like with gas and gauges.
    • When justification to believe does not comes from justification to believe other propositions, this justification is immediate.
  • Some clarifications:
    1. The question of whether justificaiton to believe is mediate or immediate is a question of what kind of epistemic support you have.
      • It is not a question of how much support you have.
      • It is not about what physiological process you’ve undergone.
      • Support for believe can be inferential even if you did not arrive at a proposition by deriving or infering from other beliefs.
    2. In order for you have to have immediate justification, it is required that your justification comes from nowhere.
    3. The fact you have immediate justification to belief something does not ential that no other beliefs are required to form or entertain that belief.
    4. Justification is usually defeasible.
    5. Beliefs epistemically overdetermined.

Why Believe in Immediate Justification?

  • The most famous argument for immediate justification is the Regress Argument.
    • There is dialectical regrees.
    • There is grounding regress.
    • There is justification-making regress.
      • This is what we will focus on.
  • The argument is that there are four ways for the regress to play out:
    1. The regress never ends, and justification chains go on endlessly. (Infinite regression.)
    2. What makes you justified in believing P is having justification to believe other things, and what makes you believe some of those is your justification to believe P. (Closed loop.)
    3. Eventually we get a proposition believed inappropriately. (Skepticism.)
    4. Eventually we get a proposition you have justification to believe, but that justification does not come from your believing, or having justification to believe, any further propositions. (Immediate justification, knowledge, etc..)
  • Foundationalists argue that (1) and (2) are untenable.
    • We must accept (3) or (4).
Examples of Immediate Justification

I am imagine my grandmother. The way I am imagining her is sitting in her kitchen. Or at least, I believe it is. And it seems I could be justified in that belief.

I think about a domino and a chessboard. It is obvious to me that the only way to wholly cover two spaces on the board is to place the domino horizontally or vertically.

June 11th, 2013 Lecture

Highlights from Last Time Handout

The AAA Scheme

  • Recall the archer, aiming, taking a shot.
    • Was it accurate?
    • Was the archer skilled?
    • If the archer hit the bull’s-eye, or was it something else?
  • Cases:
    1. Accuracy with no skill.
      • No competence
      • Right by luck
      • True belief without justification
    2. Archer uses her skill but there is a wind.
      • Competence, used the skill
      • Didn’t succeed.
      • Justified false belief
    3. Hit’s the bull’s-eye and uses some skill, but the explanation is not because of the bull’s eye.
      • Someone takes a shot at a target.
      • Gust of wind swipes it away.
      • Gust of wind swipes it back.
      • It was only the second gust of wind, not manifestation of skill because of bull’s-eye.
      • These are Gettier cases.
      • Form a true belief and you are justified, but it is not because of your competence.

Virtue epistemology, Sosa-style

  • Safety is bad
    • The archer walks to the shooting range,
    • All targets but one are covered by a force-field but one.
    • The archer finds the one without it to try and hit it.
    • The reason why their hitting is their competence.
    • So easily could they have hit the other ones.
    • In some cases of Gettier cases, like the fake barns, is like the force-fields.
    • The person does know that they are looking at barn, but so easily could they of had another belief they have an inaccurate meta-belief.
    • Even if they are getting right, it is not explained by this higher order competence.

The Pyrrhonian Problem

  • An extreme version of skepticism.
  • Self-debasing, because the belief shows the belief was incorrectly held.
  • But the way philosophy works now is you have to write an article.
  • If you reject it, why is it that it doesn’t work?
  • The response is foundationalism.
    1. Some stopping isn’t arbitray. Some beliefs are non-inferentially justified or basic.
      • The authority of experience, they don’t need to be justified themselves to justify beliefs.
      • We’re not going to worry about arbitrariness because it would be “senseless” to try and justify experiences.
      • Current experiences that you’re having right now.
      • What is your justification for believing this.
      • For now thought, we stay here, with the first question.

Infinitism

  • The structure of justification can be infinite and non-repeating.
  • It avoids skepticism because being justified in believing implies that there is an infinite, non-repeating justification for belief.

Foundationalism

  • There is a broad literature on how something can be non-inferential justification. Ways:
    • Infallibility: Belief is non-inferentially justified only if believe implies truth.
      • The cogito
      • All necessary truths (unlikely)
    • Indubitability: Belief that is non-inferentially justified only if the belief cannot be doubted.
      • When psychological, when from the agent, they can doubt their own or a crazy person can doubt all.

Problems for Acquitance Foundatinalism

  • A big problem for acquitance foundationalism is the speckeled hen.
  • Fumerton believes that you must think that you have a priori probabilistic account of justification, or skepticism!?

Pryor’s modest view

  • Does not commit you to:
    1. It isn’t infallible justification.
    2. It isn’t indefeasible justification.
    3. It isn’t what he calls “autonomous” justification. 
      +--------------------------+               +--------------+
| I'm not a brain in a vat | <-------------| I have hands |
+--------------------------+   immediately +--------------+
                                justified          ^
                                                   |
                                           +--------------+
                                           | Visual       |
                                           | experience   |
                                           | of hands     |
                                           +--------------+
  • There are worries about:
    1. Mathematical knowledge
    2. Stored knowledge, semantic memory, source amnesia

June 13th, 2013 Lecture

Coherentism

There’s no ultimate anchor for justification of belief, beliefs get individual justification by being in a system that is coherent.

  • There are different reasons for rejecting them:

Input

  • Coherentism fails because it cannot account for the rational constraints by experience.
  • Some set of beliefs could be perfectly coherent while exhibiting gross negligence with regards to reality.

Isolation and Alternative systems

  • Accordsing to the isolation objection, there is no reason to believe that sheer coherence as such is truth conducive.
  • You could have an infinite number of equally coherent systems which come up with different results about truth.

Virtuous Inconsistency

  1. It is realistic to suppose that:
    1. I’ve formed some false belief.
    2. Some of them are false. You can’t know which, and so cannot be rationally required to abandon any right.
  2. This hols even for inquirers who are totally scrupulous.
  3. If so, it is clear that all my beliefs could be justified while I am also justified in believing something.
  4. But that statement is is incosisntent with the conjuction of all my beliefs.
  5. So it can be more rational to become less coherent.

Incosistency can be rational, but truth is not consistent.

Mentalism

  • Doxastic attitudes:
    1. Belief
    2. Disbelief
    3. Suspension of judgement
  • There are two categories:
    1. Doxastic mentalism
      • Replaces “mental states” with “doxastic attitudes”
      • Coherentists
    2. Non-Doxastic Mentalism

June 18th, 2013 Lecture

  • We can imagine world in which wishful thinking is reliable, but we are still included to think that beliefs just by wishful thinking in those worlds are unjustified.
  • Another problem for reliablism is when a subject has a reliable method of justification that they don’t know is justified.
  • Your belief could be hooked up to the world in the right way.
  • Blindsighters are a phenomenon where people who cannot see still reliably “guess” correctly when they are forced to “see” something, or asked.

The Generality Problem

  • What type of process was being used?
    • Reliabalism needs to give us a consistent response to “was he justified?”
  • Depending of which type of process we choose as the relevant one to test for reliability, process reliablism yields different answers as to whether the belief is justified. Examples:
    • The process of forming a belief from perception.
    • The process of forming a beliefs about creatures on the basis of visual perception in bad lighting conditions.
    • The process of forming a belief that there
  • It’s a problem because there are many true types of reliable processes which are not always true.

Problem of Logical/Probabilistic Relations

  • The next problem is the problem of logical relations
    • If you use a totally invalid inference, you can’t come to know.
    • Whether you infer correctly is related to whether you are justified.
    • Logical and probelistics relations are not mental states, but if weak internalism says only mental states are important, it cannot explain why logical relations.

June 20th, 2013 Take-Home Midterm

Instructions

Answer prompt (1) and three other prompts of your choosing. Each answer should be one page long at minimum (double-spaced). You do not get significant extra credit just for writing a lot more than one page each. What matters is that your answer is complete, accurate, and adroit. When you give counterexamples, please indent them in their entirety and give them names. Type your answers up and print them.

Prompts

  1. Explain what the Gettier problem is and what kind of analysis of knowledge it refutes. Invent two Gettier cases of your own in explaining the problem, one of which is explained by the “No False Grounds” (NFG) view, and one of which is not obviously explained by that view. Then, explain how Gilbert Harman would try to address your second case with his more sophisticated NFG view. Having done so, evaluate Harman’s view, explaining why you think it fails or could succeed.
  2. Explain how defeasibility theorists tried to solve the Gettier problem, with details about (a) how these theorists typically defined objective defeaters, and (b) why there are objective defeaters in Gettier cases given that definition. Explain why the usual definition of an objective defeater invites the objection from “defeater-defeaters”. In doing so, give an example of your own where there is both (i) an unpossessed defeater and (ii) an unpossessed defeater-defeater given the typical definition. Explain why the defeasibility theory would wrongly predict that the subject does not know. Finally, explain whether you think the defeasibility theorist could avoid this objection.
  3. Explain Goldman (1967)’s causal theory of knowing. In doing so, explain how and why Goldman’s theory is not as simple as the following theory:

    (Too Simple) S knows that P iff S’s belief that P is caused by the fact that P.

    Having explained Goldman’s theory, offer three objections to it. If your objections involve counterexamples, invent at least one new counterexample. Finally, consider and refute a possible response on Goldman’s behalf to one of your objections.

  4. Explain Nozick’s idea that a belief must “track the truth” to constitute knowledge, with a focus on what we called the sensitivity condition. In doing so, try to explain why Nozick’s theory is initially appealing and intuitive. Explain how and why Nozick introduces the idea of a method of belief-formation in articulating his final account. What counterexample is he trying to avoid? Is the appeal to methods totally ad hoc, or is there a good independent reason for it? Finally, give two objections to the sensitivity condition. If you give counterexamples, invent one new counterexample.
  5. Explain Sosa (1999)’s idea that a belief must be “safe” to constitute knowledge. Why is Sosa’s safety condition not equivalent to Nozick’s sensitivity condition? List some advantages that safety has over sensitivity. Why is the distinction between safety and sensitivity tied in with a discussion of skepticism? How does Sosa try to explain away the appeal of the skeptic’s premise that you can’t know you are not a handless brain in a vat? End with one objection to safety as a necessary condition for knowledge.
  6. Explain Sosa (2007)’s “triple-A” evaluation scheme, using examples of a kind of performance other than archery. Then explain how the differences between
    1. animal knowledge,
    2. merely justified true belief,
    3. justified false belief, and
    4. mere true belief

    are parallel to examples with the type of performance you chose. Explain why Sosa denies that apt performances must be safe. Why does Sosa’s distinction between animal knowledge and reflective knowledge become important once this is recognized? List two major attractions of Sosa’s view. If you think the view doesn’t ultimately succeed, explain why. If you think that it does succeed, explain why you are unconvinced by one of the objections we discussed in class.

  7. Explain what foundationalism says in its most generic form, and how foundationalists try to resolve the Pyrrhonian Problematic. Then, explain two specific versions of foundationalism (e.g., infallibility foundationalism, acquaintance foundationalism, Pryor’s view), giving two objections to each. Explain how one might construct a revised version of foundationalism that avoids these objections. At the end, explain whether and why you think the generic foundationalist thesis is true or false.

Reponse

Question 1

Justified True Belief

Since Aristotle, the literature of epistemology has been centered on the Justified True Belief (JTB) theory of knowledge. The theory being that a person S knows that P if and only if

  1. P is true.
  2. S believes that P, and
  3. S is justified in believing that P.

These have been speculated to be both necessary and sufficient conditions, meaning that if S lacks any of them, S does not know, and (supposedly) if S has all of them, in every case S knows. A prime example is a mathematical proof, where S is presented with the demonstration of P, comes to believe P on the grounds of the valid proof, and the proof is of the Pythagorean theorem, which is true. This fits really well with intuition, this gloss feels clean, light, like a suitable definition.

Gettier’s Counterexample

In his 1963 paper, “Is Justified True Belief Knowledge?”, Edmund Gettier ingeniously comes to what he thinks is a counterexample to the (JTB) model. Gettier has the reader not that for the proposition P, where it entails Q, if S is justified in believing P, then S is also justified in believing Q. This is perfectly non-offensive and obvious, but Gettier shows that this can lead to S justifiable believing something that is in fact false. These examples can take two logical forms, abusing the “and” and “or” operators.

Imagine that S goes out to the bar where he meets D and F. F is quite the drinker, has been taking shots the whole time, and is behaving noticeably intoxicated. D is quite the talker, and tells S that she is a DJ. She shows S pictures of DJ gear, tells vivid stories, and reliably answers questions about the vocation. S comes to justifiably believe that D is a DJ. S has just taken an undergraduate class in logic, and eager to use that new knowledge, notes that “D is a DJ or F is sober” is a true proposition because of S‘s belief that D is a DJ. As it turns out, D is a huge liar and self-obsessed. Furthermore, F just likes the excuse of intoxication so that socializing is easier, and has an agreement with the barman about only being served water. Notice that S believed justifiably a true proposition, that “D is a DJ or F is sober,” and that it is tempting to say that S still does not know. So something is missing from (JTB)!

Harman’s Response

In Thought, Gilbert Harman believes it is easy to respond to this sort or counterexample to (JTB). Harman sees that the problem with Gettier cases is that S has false grounds for justifiably believing a true proposition. So, reasonably, Harman notes that if a “no false grounds” clause (NFG) was added as another necessary and sufficient condition to (JTB), these Gettier cases could be accounted for. More formally, S knows that P if and only if all conditions of (JTB) and all of S‘s grounds for believing P are true. Unfortunately (JTB + NFG) does not work because it is equally susceptible to Gettier cases.

Imagine that S is still at the bar when he gets a call from his friend. His friend, G, is just recently single, and is looking to get back into the dating world. G asks S what the party is like, and S is looking at someone, H, who is just G‘s type, so S comments that the party is just the type that G wants to be at. S is a really good judge of character and knows G really well, so assume he has impeccable visual and other evidence of their going well together. Unbeknownst to S, it is actually LGBT night at this particular bar, and it so happens that G is heterosexual. But it also so happens that S was right about H being a G‘s type, but H was at the bar with a LGBT friend. Here, S does not reason from anything false, but he does not know because had he turned his attention to any other person he would have noticed that it was LGBT night and that G would not enjoy it.

Question 3

Goldman’s theory

In his “A Causal Theory of Knowing,” Alvin Goldman takes a specific Gettier case, analyzes it, and generalizes it to a new take on the old “S knows that P.” He begins with the case where S has good grounds for believing that

(Q) Jones owns a Ford.

As a logic game, S notices that it is true that

(P) Either Jones owns a Ford or Brown is in Barcelona.

As in all Gettier cases, this belief is a (JTB), and it also so happens that Jones does not own a Ford by Brown is in Barcelona. What Goldman wants to say about this is that what makes (P) true is that fact that Brown is in Barcelona, but the fact that Brown is in Barcelona has nothing to do with S believing that (P). The requirement of a causal connection is what Goldman wants to add to (JTB). Goldman posits that there has to be a proper sort of causal chain between belief and justification. Formally, Goldman’s theory is as follows,

(JTB+ACC) S knows that P if and only if the fact P is causally connected in an appropriate way with S‘s believing P.

This means that S‘s belief fails because the causal connection is lacking, and this is the right outcome.

Counterexamples

It is possible to explain Goldman theory with this (as we shall see) too simple formulation, which is:

(Too Simple) S knows that P if and only if S‘s belief that P is caused by the fact that P.

In order to understand how this is not the same as (JTB+ACC), imagine a case where S‘s belief that P is cause by P‘s truth, but S does not know because the chain is inappropriate. For instance, imagine that S is very, very religious and has a book that he will believe whatever is found inside. G is a mathematician and really respects S‘s religiosity. G decides to express this respect by, unbeknownst to S, writing in S‘s book the conclusions of many mathematical proofs. When S comes to read them, being very religious, accepts them unquestionably. In this case, (Too Simple) fails because S‘s belief that P is actually caused by the fact that P, which G discovered and shared. But still, if G had written down the wrong formulas, S would have still believed, the intuition is that S did not know. While some may think that this is chain of events is also a counterexample to (JTB+ACC), Goldman would say that this is not an “appropriate” way for S‘s believing. In this way, (JTB+ACC) gets a different result to (Too Simple), and so is distinct.

This pattern of counterexample can be generalized. The way that it works is by the fact that the belief is in some sense causally connected with the truth of the belief, but it somehow not plugged into the world in the right way. Alternatively, that the subject was only luckily holding the right belief through this chain.

A classic manifestation of this pattern is the papiere-mache barn case. It is as follows,

Henry drives into a part of the state that he’s never visited before. He is looking at a real barn, and has impeccable visual and other evidence that it is a barn. His justification is sound in every way. However, in the rest of this unusual part of the state there only fake, papiere-mache barns, any of which would have fooled Henry into thinking it was a barn.

The explanation of this is that Henry sees a barn and forms not inferential beliefs about it, he just accepts that what he sees is a barn. This belief is true and justified. The reason it is a problem for both (JTB) and more specifically (JTB+ACC) is that Henry does not rely on a false inference. The causal process in this case is one of perception, which Goldman notes is a highly reliable one (along with memory). Yet this causal process which is hooked up to the world and the truth in the right way still falls short of knowledge, so it is a counterexample to (JTB+ACC).

Another terrific (or devastating …) type of counterexample to Goldman’s (JTB+ACC) are ones where the subject is given true beliefs about the world, “veritable,” but those beliefs are caused by a third party. A common way of instantiating it is to imagine a “boring neuroscientist” who causes in a subject, who is a brain in a vat, nothing but veritable delusions. The world that this subject has access to represents the world outside of the vat exactly, and the subject comes to form justified true beliefs which a causally connected to the facts. Yet, this subject does not know these things.

Response

I think that there is still hope for the causal theory of knowledge. The failures of the theory with epistemic luck, deviant causal chains, and veritable delusions is failure to understand which chains are “appropriate.” This is not a catch-all for any instance when intuition disagrees with (JTB+ACC), but rather a challenge to discover the means by which we can acceptably come to know things. The reason I think there is still hope for appropriate causal chains is because my model of mind is a naturalistic and my model of reality is deterministic. This encourages me about a causal theory because I think that all perceptions, memories, and inferences are actually natural events within the brain. I hope that with a better understanding of the inner-workings of the brain and a better understanding of how the brain collects information from the world, a rigorous and intuitive definition of “appropriate” can be formulated.

Question 7

The Pyrrhonian Problematic

The Pyrrhonian Problematic is an extreme version of skepticism. While the view is self-debasing in the sense that accepting the view entails one’s being unable to hold the view, a formulation of it would look something like this:

  1. One must be justified in believing.
  2. There are only three ways to be justified:
    1. Infinitely based on non-repeating justification,
    2. Recursively justified (circularly), or
    3. Ending the justifications at a foundational justification.
  3. Infinite justification is impossible, unacceptable.
  4. Recursive justification is not justification.
  5. Foundational justification is arbitrary.
  6. Therefore, justification is impossible.
Generic foundationalism

The answer that the foundationalist gives to the Pyrrhonian is that some foundational justification is not arbitrary. Some beliefs are non-inferentially justified. More formally, there exist non-inferentially justified true beliefs that an epistemic agent derives all other beliefs from. Examples of beliefs that might be non-inferentially justified are ones formed immediately by perception, like Henry noticing the barn in papiere-mache barn country. Or perhaps the subject S noticing the truth of a mathematical proof is becoming immediately acquainted with truth.

As these two examples show, there are different ways that a belief could be non-inferentially justified. For instance, some foundationalists think that S‘s belief that P is non-inferentially justified if and only if S‘s believing that P entails that P is true, which is called infalliblism. This would mean that everything that is necessarily true is non-inferentially justified, which is unlikely. This is unlikely because imagine that S is presented and accepts the proof Euler’s formula. It is not tempting to say that this is non-inferentially justified belief. Furthermore, the connection between truth and belief is so strong that skepticism likely follows from this theory.

There is also “acquaintance foundationalism.” Its formal definition is:

S is non-inferentially justified in believing that P if and only if S is directly acquainted with the fact that P, her thought that P, and the relation of correspondence between the fact and the thought.

This theory suffers from an issue when S is presented with something like a “many speckled hen.” Imagine that the hen has 47 speckles. S is directly acquainted with the fact that the hen has 47 speckles when the hen comes in to her visual experience, and S has the thought (in some way) that the hen has 47 speckles. Yet, despite this, unless S is “rain man”, she likely does not know that the hen has 47 speckles, despite this acquaintance.

I think generic foundationalism is true and that the right modifications could allow it to correctly explain all of these cases. Foundationalism has to be fitted for human perception in that it correctly accounts for those perceptual qualities that human beings can reliably distinguish and those that confuse the human senses, those not included in inherent human competencies. I believe this, in part, because if the Pyrrhonian trilemma is not a false trilemma, then the only way we could come to know things was from basic, non-inferential, brute justification. And because I think we do come to know things (this is an intuition), I think the solution must be non-inferential justification. Furthermore, my model of mind and reality is a causally deterministic one, which leads me to think that there are low-level causal chains which give rise to basic concepts in the mind. These truths may be too basic for high-level human language, which is based on thought, to describe. This type of formulation would be difficult because it is hinged on a low-level understanding of the brain and how the sense-data comes to be consumed by the brain. The vocabulary needed to restrict non-inferential justification via sense data includes terms like order of magnitude and “on/off” ordered states of small clusters of neurons.

June 25th, 2013 Reading

“Is Knowledge Easy — or Impossible?” by James Van Cleve

  • Cleve has defeneded the idea that we obtain knowledge of the reliability of our faculties with the faculties,
    • People think this is circular.
  • It requires epistemic externalism.
    • Any argument that says we can have knowledge of the reliability of a faculty using the faculty makes knowledge too easy.
    • In what follows, Cleve presses the suggestion that the only alternative to such externalism is skepticism.
      • This is important because those that reject the circularity likely do not want to embrace skepticism.
      • Usually these people believe this is a false dilemma, and try for the medium between the externalist’s easy knowledge and the skeptic’s unattainable knowledge.

Varieties of Externalism

  • He is going to define internalisms positively and externalisms as the rejection of the preceding internalism.
    • The core tenet of the internalist view is:

      There is no first-order knowledge unless there is also higher-order knowledge with respect to the factors that make first-order knowledge possible.

    • Theories that fit under this slogan may vary in two dimensions:
      1. The object of higher-order knowledge
        1. That the knowledge-making or knowledge-conferring factor obtains
        2. That the factor, whatever it may be, has “justificatory efficacy.”
      2. The nature of higher-order knowledge
        1. Must the higher-order knowledge be:
          1. Actual? or
          2. Potential?
        2. Must the knowledge arise or be obtainable:
          1. In some prescribed way, typically by reflection alone
          2. In any way?
  • The externalism that will be important is that one where it’s denied that a subject need actually know that the factors justifiying his belief or giving hum knowledge or reliable indicators of truth.
  • For a reliabalist, the fact that the subject’s belief has been produced by a reliable process is enough for justification.
  • But reliabalism is not the only variety of externalism,

    The concept of epistemic justification is … internal … in that one can find out directly, by reflection, what one is justified in believin at any given time.

    • This makes him an internalist in the two senses from above:

      One is justified by some factor in believing something one must be able to know by reflection both that the factor obtains and does not confer justification on belief in that belief.

    • There are sources of justification or knowledge that deliver their goods even if the subject does not know they are reliable for Chisholm.
  • The “track-record argument”:
    1. At t1, I formed the perceptual belief that p, and p.
    2. At t2, I formed the perceptual belief that q, and q.
    3. Therefore, sense perception is a reliable source of belief.
  • There is a strong tendency to regard such arguments as circular.
    • Is that the second conject in each premise could be known only on the basis of sense perception.
  • Externalism suggests that what needs to be added is knowledge of reliability.

    (KR): A source K yields knowledge for a subject if and only if S knows that K is reliable.

(KR) as Leading to Skepticism

  • We find ourselves in the following predicament:
    1. We can know that a deliverance of K is true if and only if we first know that K is reliable.
    2. We can know that K is realiable if and only if we first know, concerning certain of its deliverances, that they are true.
  • If both are true, skepticism is the only option.

The Reidian Alternative

  • Basic first principles:
    1. The exitence of all things which I am concious.
    2. That those things I remember did really happen.
    3. That we perceive by our senses is really what we perceive.
    4. The natural faculties we distinguish truth from error are not fallacious.

A formulation of my own

Description

  Truth +----------------------------------------------+ 
    |   |               +----------------------------+ |       
    |   |          (a)  |         (b)   +----------+ | |  (c)
    +-----Events ------>| Senses ------>| Thoughts |-----------+
    |   |               |               +----------+ | |       |
    |   |               +----------------------------+ |       |
    |   +----------------------------------------------+       |
    |                                                          |
    +----------------------------------------------------------+
  • The set of thoughts is contained within the set of senses, and the set of senses are contained within the set of events.
    • Everything natural is in the set of events.
    • Thoughts are events.
  • The process of (a) is one in which a conditional event, an event conditional on other events, comes to be triggered.
    • The event of coming to sense a table is conditional on their being a table in that the lack of the table would result in the lack of the sense-event.
  • The process of (b) is one in which sense events are aggregated and synthesized by our mind.
    • The actual table triggers the mental event of sensing it, which triggers the thoughts about the table.
  • There are events which occur external to our mind.
    • They have properties, spatial, temporal, and natural.
    • These properties are abstractable to “information.”
  • Senses causally detect at least some of the events and at least some of the information.
    • Our senses are limited in that they capture a subset of the available information about events.
  • Our thoughts have access to the information accrued from the senses.
    • This forms an internal model of the external world.
  • Thoughts and thinking are a complex, specific case of sense, perception, which are complex, conditional events. They are the aggregation of many basic sensory perceptions, and the ordering and synthesis of perception.
    • Thinking and reasoning are a form of sensing that do not have a way of directly interacting with events.
  • Senses are a type of event which is about another event, events which rely on other events for their content.

Explanatory power

  • Knowledge is when the thought-event is of an actual event which came through sense-events.
    • Gettier cases are not knowledge because the actual event isn’t “actual” and it does not come through the senses, the “knowledge” occurs without the appropriate sense, it is a sense-independant thought which does not represent an event.
  • We have evolved to distinguish relevant information from events, and perception is the process of taking events as input (a) and sending a translation to the language of thought to the higher faculties (b).
    • Evolution is the process of increasing the complexity of a system given energy, becoming increasingly fit in being better at reproducing. Because we’re here and a far greater number of possible beings aren’t, we have good reason to believe our senses are more apt at distinguishing events which make repoduction more likely.

“Basic Knowledge and the Problem of Easy Knowledge.” by Stewart Cohen

Basic Knowledge and the KR Principle

  • The problem is that we can’t use anything but our senses to determine if our senses are reliable except our senses. Bootstrapping.

    (KR): A potential knowledge source K can yield knowledge for S, only if S knows K is reliable.

  • An evidentialist is someone who denies one can know P on the basis of evidence E without knowing that E is a reliable indication of P

The Problem of Easy Knowledge: Closure

If S knows P and S knows P entails Q, then S knows (or at least is in a position to know) Q.

  • Nozick uses his view to deny closure.

If S knows P on the basis of R, and D is a defeater of R as a reason to believe P, then S knows D is false.

June 27th, 2013 Lecture: Testimony and Justification

  • There are various pathways to knowledge.
    • Some pathways are generative
      • Perception, for example.
    • Some pathways are preservative
      • Inference
  • There are controversies about how to classify certain sources of knowledge. – You may think that they are preservative in that the follow way:
    > It’s natural to think that what memory does is transmit > some peice of knoweldge or justified belief from one’s > earlier self to one’s later self.
  • A strong version of preservative view of testimony is that in order to come know P from testimony, the testifier must know P.
  • Lackey thinks that people who don’t know something can come to have you know that P. > JP: If S reports that P to H and H has no defeaters for > S‘s report that P, then H is justified in accepting that > P on the basis of S‘s testimony.

July 4th, 2013 Paper

Instructions

Write a paper of 6–10 pages (double-spaced) on a topic below. You are more than welcome to invent your own topic. Just create a prompt and pass it by me first.

Topics

  1. Which of the approaches to the Gettier problem gets closest to solving it, in your view? Explain the solution and compare and contrast it with some others, making clear why it fares better. Then anticipate and answer some objections to it. If some supposed Gettier cases are left unexplained, explain why we shouldn’t take them seriously.
  2. To know that P, do you need to be in a position to know that you know that P? If you answer ‘yes’, defend your answer with at least three arguments, and consider and rebut at least two objections. If you answer ‘no’, give a theoretical argument for your conclusion, and also give some examples where you think it is clear that someone knows that P without being in a position to know whether she knows that P.
  3. Is all circularity vicious? If ‘no’, what is the difference between epistemically vicious and epistemically benign circularity?
  4. Is a brain in a vat with all the same experiences, apparent memories, and other internal states as you really equally justified in believing everything as you? If you think there is a difference in justification, what accounts for it, and how can you explain intuitions to the contrary? If you think there is no difference, why care about justification? Wouldn’t this show that justification is irrelevant to the goal of truth?
  5. Can the generality problem for process reliabilism be solved?
  6. Can a more sophisticated form of coherentism avoid the objections we discussed in class? (Feel free to focus on a subset of those objections to add focus to the paper.)
  7. Cohen and others claim that every view that allows for one to gain knowledge via a source without antecedently knowing that the source is reliable will legitimate some form of vicious bootstrapping (a la Vogel’s Roxanne case). Explain the argument in favor of this claim. Is this claim true? If true, what should we make of it? If false, what is wrong with the argument?
  8. Can perceptual seemings (and other sorts of seemings, such as apparent memories) provide some justification all on their own? If so, what is the relevant difference between a perceptual seeming and a belief? If not, is this because seemings are too much like beliefs? (If so, how are seemings like beliefs? How are they evaluable?)
  9. Can modest internalists like Pryor avoid the problem of the speckled hen?
  10. Is Sosa’s distinction between animal knowledge and reflective knowledge a real distinction? Is it really plausible that we have the former in Fake Barn Country?
  11. What is the best argument for skepticism, and where does it go wrong?